© The Institution of Engineering and Technology
The need to support various recognition applications on energy-constrained mobile computing devices has steadily grown. Exploiting common characteristics of recognition algorithms, a very energy-efficient multiplier that can support a runtime trade-off between computational accuracy and energy consumption is proposed. Compared to a precise multiplier, the proposed multiplier consumes 11.6×–3.2× less energy per multiplication while achieving 82–99% of the computational accuracy with negligible negative impact on recognition accuracy for three different recognition applications.
References
-
-
1)
-
4. LeCun, Y., Cortes, C., Burges, C.: ‘The MNIST database of handwritten digits’ .
-
2)
-
3. Babića, Z., Aramović, A., Bolić, P.: ‘An iterative logarithmic multiplier’, Microprocess. Microsyst., 2011, 35, (1), pp. 23–33 (doi: 10.1016/j.micpro.2010.07.001).
-
3)
-
6. Liberman, M., Amsler, R., Church, K., et al: ‘TI 46 words speech database’ .
-
4)
-
1. Hegde, R., Shanbhag, N.: ‘Energy-efficient signal processing via algorithmic noise-tolerance’. Int. Symp. Low Power Electronics and Design, San Diego, CA, USA, August 1999.
-
5)
-
2. Kulkarni, P., Gupta, P., Ercegovac, M.: ‘Trading accuracy for power with an underdesigned multiplier architecture’. Int. Conf. VLSI Design, Chennai, India, January 2011.
-
6)
-
7. Sakhi, O.: ‘Face detection using support vector machine (SVM)’ .
-
7)
-
5. Reservoir Lab: ‘Reservoir computing’ .
http://iet.metastore.ingenta.com/content/journals/10.1049/el.2013.4212
Related content
content/journals/10.1049/el.2013.4212
pub_keyword,iet_inspecKeyword,pub_concept
6
6